Method and device for optimizing the compression of a video stream

Abstract

A method and device for compressing at least a portion of a video image for transmission in a video stream such that the compressed portion either includes or does not include an encoded residual depending on whether including the residual would be advantageous in terms of data rate and distortion. The size of the video stream and the resources needed for decompression at a receiver are optimized through encoding only the most relevant data in the video stream that is to be transmitted. If the information provided by a residual is insufficiently relevant, the residual is either not encoded or is not included in the transmitted video stream. Factors associated with a residual are only encoded and transmitted if use of the residual would limit the distortion in a satisfactory manner and if the encoding of the associated factors does not generate too high of a data rate. The disclosed method and device make it possible to minimize and optimize resources needed by a decoder by a receiver of the transmitted video stream.

The invention claimed is: 
     
       1. A video data compression apparatus having one or more hardware processor having ability to compress a portion of a video image, comprising:
 circuitry using at least one said processor configured to determine a prediction of said image portion, wherein said prediction may have an associated residual; 
 circuitry using at least one said processor configured to determine a distortion associated with the prediction; and 
 circuitry using at least one said processor configured to decide whether to encode the prediction alone or to encode both the prediction and an associated residual, the decision being based at least upon a determined distortion associated with the prediction; 
 wherein the decision of whether to encode the prediction alone or to encode both the prediction and an associated residual includes calculating a first rate-distortion based parameter and calculating a second rate-distortion based parameter. 
 
     
     
       2. The video data compression apparatus according to  claim 1  further comprising: circuitry using at least one said processor configured to determine a residual;
 circuitry using at least one said processor configured to determine a distortion associated with the residual; and wherein the decision of whether or not to encode the residual is also based upon the distortion associated with a residual. 
 
     
     
       3. The video data compression apparatus according to  claim 1 , wherein the decision of whether to encode the prediction alone or to encode both the prediction and an associated residual further includes determining whether the second rate-distortion based parameter is greater than the first rate-distortion based parameter. 
     
     
       4. A video compression and transmission method for use by a hardware platform or a system having an ability to compress and transmit a portion of a video image, the method executed at least in part by using one or more computer processor configured to perform operations comprising:
 determining, using at least one said processor, at least one prediction of said image portion wherein said prediction may have an associated residual; 
 determining, using at least one said processor, a distortion associated with a prediction; 
 determining, using at least one said processor, a residual associated with said prediction; and 
 deciding, using at least one said processor, whether to encode the prediction alone or to encode both the prediction and an associated residual, the decision being based upon at least two decision parameters and includes calculating a first rate-distortion based decision parameter and calculating a second rate-distortion based decision parameter wherein at least one decision parameter is calculated based upon the determined distortion associated with the prediction. 
 
     
     
       5. A method using at least one processor according to  claim 4 , further comprising:
 determining a rate associated with the residual; 
 determining a distortion associated with the residual; and 
 wherein the deciding is also further based upon the determined rate associated with the residual and the determined distortion associated with the residual. 
 
     
     
       6. A method using at least one processor according to  claim 4 , further comprising:
 transforming said residual to the frequency domain; 
 quantizing the transformed residual; 
 dequantizing the quantized transformed residual; 
 transforming the transformed, quantized, and dequantized residual to the space domain; and 
 determining the rate associated with each residual being based on said residual and on the transformation of the transformed, quantized, and dequantized residual to the space domain. 
 
     
     
       7. A method using said at least one processor according to  claim 4 , wherein deciding whether to encode the prediction alone or to encode both the prediction and an associated residual further comprises comparison of the first rate-distortion based decision parameter (RD1) and the second rate-distortion based decision parameter (RD2), wherein the first and second rate-distortion based decision parameters are calculated as follows:
   RD1= k+Dp;    
   RD2= k+Rr*ε+Dr    
 Where: 
 Dp is the distortion associated with each prediction; 
 Rr is the rate associated with the residual; 
 Dr is the distortion associated with the residual; 
 ε is a positive predetermined factor; and 
 k is an optional term. 
 
     
     
       8. A method of  claim 7 , wherein after said deciding:
 (a) if RD1≦RD2, only every prediction is encoded and transmitted; and 
 (b) if RD1>RD2, every prediction and each residual is encoded and transmitted. 
 
     
     
       9. The method according to  claim 7  further comprising determining a rate, Rp, associated with every prediction, and calculating the optional term k, with k=Rp*ε, where ε is a positive predetermined factor. 
     
     
       10. A method of according to  claim 4  wherein said deciding further comprises using said at least one processor for comparing a decision parameter RD1 and a threshold value RDt, with RD1=k+Dp where:
 k is an optional term; and 
 Dp is a distortion value associated with a prediction. 
 
     
     
       11. A method of  claim 10 , wherein, if RD1 RDt, only every prediction is encoded, and if RD1>RDt, every prediction and the residual are encoded. 
     
     
       12. The video data compression apparatus according to  claim 1  wherein deciding whether to encode the prediction alone or to encode both the prediction and an associated residual further comprises a comparison of the first rate-distortion based parameter (RD1) and the second rate-distortion based parameter (RD2), wherein the first and second rate-distortion based parameters are calculated as follows:
   RD1= k+Dp;    
   RD2= k+Rr*e+Dr    
 Where: 
 Dp is the distortion associated with each prediction; 
 Rr is the rate associated with the residual; 
 Dr is the distortion associated with the residual; 
 e is a positive predetermined factor; and 
 k is an optional term. 
 
     
     
       13. A video compression and transmission method performed using a hardware platform or by a system having an ability to compress a portion of a video image, the method executed at least in part by using one or more information processor configured to perform operations comprising:
 determining at least one prediction of said image portion wherein said prediction may have an associated residual; 
 determining a distortion associated with each prediction; 
 determining a distortion associated with each residual; and 
 deciding, using at least one said processor, whether to encode the prediction alone or to encode both the prediction and an associated residual, the decision being based at least upon one or more decision parameters, wherein at least one decision parameter is calculated, using at least one said processor, based upon a determined distortion associated with the prediction and a determined distortion associated with the residual, and wherein the decision of whether to encode the prediction alone or to encode both the prediction and an associated residual includes calculating a first rate-distortion based decision parameter and calculating a second rate-distortion based decision parameter. 
 
     
     
       14. The video data compression apparatus according to  claim 13  wherein at least one decision parameter is calculated, using at least one said processor, based upon a determined distortion associated with the prediction and at least one other decision parameter is calculated, using at least one said processor, based upon a determined rate and distortion associated with the residual. 
     
     
       15. A device for compressing at least a portion of an image, comprising:
 an image information processing device having one or more hardware processors configured to determine at least one prediction of said image portion, 
 one or more processors of said processing device configured to determine a distortion value for each said prediction; and 
 one or more processors of said processing device configured to decide whether or not to encode a residual based on at least the distortion values associated with every prediction, wherein the decision of whether or not to encode the residual includes calculating a first rate-distortion based parameter and calculating a second rate-distortion based parameter.